JP3412441B2 - Image processing device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, and more particularly to an image processing apparatus having a function of detecting an input original image (document, form, original, etc.) when it is input again.

[0002]

2. Description of the Related Art With the development of information systems, the number of customers who intend to introduce a computer into their business is increasing. The customer's main desire is to save information, enter information quickly and accurately, and have a complete database.

In a conventional image processing apparatus, for example, an optical character reading apparatus, a stamp mechanism for imprinting a stamp on the read form is used to determine whether or not the form to be input has already been read. There was something to prepare.
Further, in addition to the determination as to whether or not the document has been read, there is a device provided with a numbering printing mechanism for printing information such as a serial number on the read document for the purpose of associating the document with the read information. In these cases, the operator visually judges whether the form has been read or not.

However, when the operator erroneously inputs the read form into the apparatus despite the numbering printing or stamping, the information of the same form is read twice. For example, when the optical character reading device is used for deposit processing, if the same form is read twice, the deposit may be duplicated, which may cause a serious problem in the subsequent procedure.

Therefore, in order to prevent accidental double reading, when the form is read, the printing area of the numbering or stamp is also read, and the presence or absence of the numbering or stamp is recognized from the image information. Therefore, there is an optical character reader having a function of preventing double reading. In this case, when the erroneously read form is input again to the apparatus, it can be detected that the form has already been read. However, in this method, the numbering print or stamp is optically recognized,
If there are characters, prints, color backgrounds, or ground patterns that cannot be dropped out in the numbering or stamp printing area,
There is a possibility that it cannot be correctly identified.

Therefore, this function can be used only for a form that has no characters, prints, or color backgrounds that cannot be dropped out in the numbering or stamp printing area. Double reading is not prevented, but it is not disclosed in JP-A-58-186
Reference numeral 876 is a device for printing on the postal matter as a code the recognition result of the address or the postal code on the postal matter, thereby preventing double printing by detecting the presence or absence of the code. In this example, since the optical detection means by the code reader is used, if characters or the like are written in the code print area on the mail piece, reliable detection becomes difficult.

[0007]

SUMMARY OF THE INVENTION According to the present invention, an input object is placed in an arbitrary area designated by the operator on the input object, that is, in an area on the input object where characters, prints, and ground patterns that cannot be dropped out are present. An image processing apparatus having a function of preventing double input by forming a predetermined mark on the image of FIG.

[0008]

In the image processing apparatus of the present invention, a mechanism for taking in image information of each form to be input and a character information of a predetermined area are cut out from the input image information and recognized. It has a function and a mechanism for forming (printing) a specially shaped mark (mark) indicating that the document has been read on the form after reading.

Further, when the read-out form is input again, it is not possible to drop out even if characters, figures, ground patterns, prints, etc. that cannot be dropped out overlap with the special mark in the area on the form. The special mark can be identified by using a detection logic that removes characters and the like, cuts out, detects, and identifies the special mark.

Further, when it is difficult to detect and identify by the detection logic, a plurality of image information is taken in to select other image information which can be optimally detected and identified from the plurality of image information, and the special mark is selected. Can be detected and identified.

The present invention is an image processing apparatus having a scanner function and capable of forming a mark for preventing double input in an arbitrary area designated by the operator on an input object after input, such as a facsimile apparatus and a document. In the image registration device, a mark having a special shape indicating that the input (read) form has been read is stamped or recorded by numbering in an arbitrary area (print recording area) on the form ( Recording by mark forming mechanism or printing mechanism).

Then, the print recording area is cut out from the input image (acquired form image), and it is determined whether or not there is a special mark in the print recording area (execution of detection logic).

Here, the special shape mark means a shape such as "■", but the size or shape of "■" is not particularly limited. Furthermore, the number of printed marks having a special shape is not limited to one. This is to separate or separate the mark component from the component other than the mark when there is a print component that does not drop out or a component such as a character in the read print recording area in the captured image information. It only needs to have a shape suitable for the detection logic described later.

As the detection logic, first, the captured image information, that is, the document position coordinates at the top, bottom, left and right ends of the document are detected from the image buffer. After that, the print recording area is determined based on the form position coordinates, buffered, and the black component in the buffered print recording area is thinned and thickened in the vertical and horizontal directions.

As a result, even if a black component other than the mark exists in the print recording area, the black component other than the mark can be erased from the buffer. Further, it is determined whether the black component remaining by this processing approximates the mark shape, and it is discriminated whether or not there is a mark, that is, whether or not it is a read form.
However, the method of determining the presence / absence of marks includes the determination of the number of marks or the arrangement position of the marks as well as the approximate determination of the mark shape.

Further, when it is difficult to determine the presence or absence of a mark, other image information that can be detected and identified optimally is selected from the plurality of captured form image information to detect the formation of the special mark. , And retry processing is performed so that they can be identified.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below in detail by taking an optical character reading device as an example, but the present invention is not limited to this device. An image processing device that needs to prevent double input, such as a scanner function,
The present invention is also applicable to an image processing apparatus capable of forming a double input prevention mark in an arbitrary area designated by the operator on an input object after inputting, for example, a facsimile apparatus and a document image registration apparatus. Each has the effect of preventing double transmission and double registration.

FIG. 1 shows a schematic structure of the optical character reader. The form 1 is conveyed along the paper feed direction 2 along the conveyance path 3. After that, the form 1 passes through the optical reading unit 4, and for example, due to a light source system (red light source + sensor etc.) 5 having a red light source, a ruled line or a character component having a red wavelength, that is, red on the form Capture the image information by removing the dropout color component. Further, for example, a light source system (blue light source + sensor, etc.) 6 having a blue light source is used to remove a component in which a ruled line or a character has a blue wavelength, that is, a blue dropout color component on a form to remove image information. Take in. In this way, the data read by at least one or more optical systems is developed as a video signal on the memory of the image buffer 7. Further, after the read designation character is recognized by the character recognition program based on the image information in the image buffer 7, the form 1 passes through the read record printing section 8 and prints the read print record 9 on the form 1.

FIG. 2 shows an example of the form layout. The read print record 11 recorded on the form 1 after the form is read
Is not limited to a fixed position on the form 1. On the form 1, a character designation area 10 that the user wants to read is designated in advance. The read print record 11 may be formed anywhere in the form 1 as long as it is other than the character designation area 10 on the form format.

Further, if it is not necessary to recognize the character data in the read character designating area 10, the data may be printed in the read character designating area 10. Further, the read print record 11 may be formed (printed) on the form 1 at a plurality of locations.

FIG. 3 shows the image information of the image buffer 7 in FIG. If the form with the read print record is input to the optical character reader again,
The read print record 15 exists on the form image 13 in the image buffer 7. The image processing apparatus of the present invention adds a margin to cut out (area 14, FIG. 3) and buffers based on the read print recording area 12 (coordinates, FIG. 2) in the form format designated in advance by the user. .

FIG. 4 shows the processing for the buffering area. In the buffering area, that is, the cut-out area image image 19, the read print recording component 1
In addition to 7, there may be non-dropout color components 16. That is, even if a plurality of light sources such as the light source system (red light source + sensor etc.) 5 and the light source system (blue light source + sensor etc.) 6 shown in FIG. 1 are used, printing on a form, or handwritten ruled lines and characters Sometimes, the wavelength component of the color that a symbol, etc. has cannot be removed.

In the present invention, the cut-out area image image 19 is subjected to the vertical thinning processing 20 and the vertical thickening processing 21 to generate the cut-out area image image 22 after the vertical thinning processing and the horizontal thinning processing. 23 and the horizontal thickening processing 24 are performed to generate the cut-out area image image 25 after the horizontal thinning processing, and finally it is possible to detect or extract only the read mark 18 component (FIG. 4).

FIG. 5 shows a schematic flow of a logic (detection logic) for detecting the read mark 18 from the buffer of the cut-out area image image 19 (FIG. 4). From the mark detection logic start unit 26 to the mark detection logic end unit 35, first, the form edge detection unit 27 detects a form edge from within the image buffer, and the print recording area determination unit 28 detects on the image buffer 7 shown in FIG. Printed record 15
3 is buffered, the cutout area 14 shown in FIG. 3 is buffered, and a cutout area image image 19 shown in FIG. 4 is taken out.

Next, with respect to the cut-out area image image 19, the vertical thinning processing unit 30 thins the black dot components in the vertical direction, and the vertical thickening processing unit 31 thickens the remaining black dot components in the vertical direction. . Further, after the black dot component remaining in the horizontal thinning processing unit 32 is thinned in the horizontal direction, the black dot component remaining in the horizontal thinning processing unit 33 is thickened in the horizontal direction. By this series of processes, the mark shape comparison and determination unit 34 determines the remaining black dot component and the actual mark shape, so that the presence or absence of the read mark 18 shown in FIG. 4 can be detected.

Here, the concept of the thinning and thickening processing is shown in FIG. For example, a case will be described where the black dots of the original image 36 are subjected to the vertical thinning processing 37. First, as the vertical direction thinning process, a process of subtracting one dot from both ends of the black line component connecting the black dot start point and the black dot end point in the vertical direction and considering the rest as a black dot component is performed. As a result, the dots 42 in the black circles remain (the dots 41 disappear).

Further, as the thickening processing in the vertical direction, 1 is applied to both ends of the black line component connecting the black dot start point and the black dot end point in the vertical direction.
A process is performed in which the dot-added component is regarded as a black dot component.
As a result, the dots 43 and the dots 44 (three pixels) are generated (the image 38 after the vertical thinning processing).

The black dots in the image 38 are subjected to the horizontal thinning processing 39 in the horizontal direction by the same method as the first vertical thinning processing. As a result, the black dot components shown in the image 40 are obtained. Is obtained. Here, for example, if the black dot component of the image 40 after the horizontal thinning process is set as the read mark 18 shown in FIG. 4, the background noise of the black dot component of the original image 36 is removed by a series of processes. hand,
The read mark can be extracted as the processed image 40.
From this process, “■” or “□” is convenient as the special mark, but another type of special mark may be provided in consideration of the background noise.

Although the black dot component has been described above, strictly speaking, from the viewpoint of the negative logic to the positive logic, the same processing can be performed with the white dot component. For pixel components,
The degree of thinning processing (strength of processing) can be appropriately determined. That is, it depends on the size and shape of the read mark 18 (special mark) shown in FIG. In reality, even if the thinning and thickening process is too strong and the special mark is faint after the process, if the mark shape comparison and determination unit 34 (FIG. 5) can detect the mark, the strength of the process is Permissible.
Further, in the mark shape comparison / determination unit 34 in FIG. 5, the number of marks, the arrangement position of marks, and the like are also taken into consideration as the determination material together with the mark shape approximation.

In the mark detection logic of the present invention, as a measure for improving the detection accuracy when the print of the mark is faint,
In addition to the thinning and thickening processing sequences shown in FIGS. 4 and 5,
It is also effective to perform thickening processing first. Further, when it is difficult to determine the presence or absence of a mark, a plurality of optical systems such as an optical system (red light source + sensor etc.) 5 and an optical system (blue light source + sensor etc.) 6 of the optical reading unit 4 shown in FIG. By selecting another image buffer 7 from a plurality of input images (group of image buffers 7) and performing the series of processes described above, by performing two or more mark detection processes (retry process), detection accuracy can be improved. It is also possible to improve.

Here, in the mark detection logic of the present invention, when it is difficult to finally determine the presence or absence of a mark, firstly, the mark is placed on a black component having a size that includes the mark. When recorded, secondly, when the read print recording area of the form is converted into black and white binary data by the optical reading unit 4 shown in FIG. Examples include the case where the mark is captured and the mark is recorded on the black component.

In the mark detection logic of the present invention, if such detection and identification are not possible, it is treated as indeterminate mark presence or absence. This indeterminate information can be finally output to the user interface screen and left to the user's judgment, or other double reading preventing means may be used in combination based on the indeterminate information. Furthermore, the second problem is solved by introducing the detection logic of the present invention in an optical character reader capable of handling multivalued data.

The case where the mark shape comparison / determination unit shown in FIG. 5 determines that the mark is indeterminate is as described above.
This is a case where the black dot component remaining after the series of processes performed from the vertical narrowing processing unit 30 to the horizontal thickening processing unit 33 is larger than the actual mark shape, and the shape approximation degree is low.

On the other hand, the case where it is judged that there is a mark is a case where the remaining black dot component is close to the actual mark shape and the shape approximation degree is high. Further, the case where it is determined that there is no mark is a case where the remaining black dot component is smaller than the actual mark shape, or there is no remaining black dot component, and the shape approximation degree is low.

The case where the shape approximation is high means that the calculated approximation value is within the margin of the approximation judgment value when the mark shape and size are compared with the actual mark shape and size. Say

The read mark in the present invention is as shown in FIG.
In this case, as indicated by ■, the size and shape are not limited as described above. For example, a shape such as ● or ★ can be used in the process of mark detection logic as shown in FIG. Even when the dropout color component 16 and the read print recording component 17 overlap each other, a good figure is shown as long as the mark can be detected. Further, the read printed record in the present invention is as shown in FIG.
7 shows a record in which the mark may be printed or only the mark may be printed together with a record such as a numbering or a stamp indicating that it has been read in the read print recording area 12 shown in FIG. As described above, when the special mark exists, the image processing apparatus can determine that the form 1 shown in FIG. 1 is a form that has been read once. As a result, it is possible to remove the restriction imposed on the form "when the double reading prevention function is used, the non-dropout color component does not exist in the read print recording area".

In the image processing apparatus according to the present invention, if a part of the input object (form) is corrected and input again (if the read print record is erased by a correction liquid or the like), double reading is prevented. The function is canceled.

[0038]

As described above, since the image processing apparatus according to the present invention forms the special mark indicating that the input has been completed on the input object that has been input once, the optical mark is formed in the area where the special mark is formed. Double input can be prevented even if there are characters, figures, symbols, and ground patterns that cannot be removed depending on the system. As a result, it is convenient to save information, enter information quickly and accurately, and have a complete database.
An image processing device can be configured.

[Brief description of drawings]

FIG. 1 is a hardware schematic configuration diagram of an optical character reading device according to an embodiment of the present invention.

FIG. 2 is a diagram showing a layout example of a form used in the optical character reading device of FIG.

FIG. 3 is a conceptual diagram showing a state of an image buffer after reading a read form in the apparatus of FIG.

4 is a conceptual diagram showing a state of a read mark detection logic in the apparatus of FIG.

5 is a diagram showing a schematic sequence of read mark detection logic in the apparatus of FIG.

FIG. 6 is a conceptual diagram for explaining a mark thinning process.

[Explanation of symbols]

1: Form, 2: Paper feed direction, 3: Conveyance path,
4: Optical reading unit, 5: Read recording print unit, 6: Image buffer, 7: Read print record, 8: Form image, 9: Cutout area, 10: Read print record on image buffer, 11: No drop Out-color component 12: read print recording component, 13: read special mark, 14: cut-out area image image, 15:
Vertical thinning processing, 16: Vertical thickening processing, 17: Cutout area image image after vertical thinning processing, 18: Horizontal thinning processing,
19: horizontal thickening processing, 20: cutout area image image after horizontal thinning processing, 21: mark detection logic start portion, 22: form edge detection portion, 23: print recording area determination portion, 24: buffering portion, 25: Vertical thinning processing unit, 26: Vertical thickening processing unit, 27: Horizontal thinning processing unit, 28: Horizontal thickening processing unit, 29: Mark shape comparison determination unit, 30: Mark detection logic end unit, 31: Original image, 32: Vertical Direction thinning processing, 33: Image after vertical thinning processing, 34: Horizontal thinning processing, 3
5: image after lateral thinning processing, 36: dots erased by thinning processing, 37: dots remaining after thinning processing, 38: dots generated by thickening processing.

Front page continuation (72) Inventor Takeyuki Sugimoto 2880 Kokuzu, Odawara, Kanagawa Hitachi Storage Systems Business Department (56) References Japanese Patent Publication 7-7434 (JP, B2) (58) Fields investigated (58) Int.Cl. ⁷ , DB name) G06F 19/00

Claims (6)

(57) [Claims] [Claim 1 further comprising an optical reading unit, the input target image input by the optical reading unit, the image processing apparatus capable of recognizing processing <br/> management of the input object image, the input target image If a special mark indicating that the input target image has been input is overlapped with the read recording area of the input target image and the input target image is re-input to the optical reading unit.
The input recording image, the read recording area is cut out.
Then, the read recording area is thinned and
By performing the thickening process, the special mark can be identified and
The feature is that it prevents the re-input of the input target image.
Image processing device to be collected. 2. The image processing apparatus according to claim 1, wherein the thinning process and the thickening process are performed after performing the thinning process on the read recording area. Image processing device. 3. The image processing apparatus according to claim 1, wherein the thinning process and the thickening process are performed after performing the thickening process on the read recording area. Image processing device. 4. The image processing apparatus according to claim 1, wherein the thinning processing and the thickening processing are thinning processing and thickening processing in a vertical direction. 5. The image processing apparatus according to claim 1, wherein the thinning processing and the thickening processing are thinning processing and thickening processing in a horizontal direction. 6. The image processing apparatus according to claim 1, wherein the read recording area includes a printing unit that prints a recording of numbering or a stamp together with the special mark. Processing equipment.